The goal is to develop coherent experimental protocols and predictive mathematical models for understanding the biomechanical interaction between myeloma-initiating cells (MICs, also known as myeloma stem cells) and bone marrow stromal cells (BMSCs, also known as bone marrow derived mesenchymal stem cells) regulating the MIC evolution. To do so, we will study the biophysical properties and signaling pathways of cancer stem cell niches (microenvironments) with the ultimate goal of developing MM cancer growth models to predict new therapeutic strategies targeting the niches. Even though multiple myeloma (MM) patients may reach a complete remission initially with therapeutic agents currently available, most MM patients eventually developed relapsed disease. Studies have suggested the presence of a small population of MICs in these patients that possess clonogenic potential and high resistance to drugs. Our preliminary studies lead to the hypotheses: 1) that MICs secret a high concentration of SDF1 which activates the SDF1/CXCR4 signaling pathway, leading to the changes in biomechanical phenotype of BMSCs and consequently, 2) that the altered BMSC mechanical properties contribute to the fate (proliferation and survival) of MICs, and thus the growth of MM. CXCR4, a G-protein coupled receptor, constitutes a control point for actin/myosin-dependent cytoskeletal signaling processes and thus regulates cell and membrane mechanics. The goals of the proposed study are to more fully characterize how the mechanical properties of myeloma BMSCs are influenced by the SDF1/CXCR4 signaling pathway, and to model and predict the impact of such changes on MIC fate by novel mathematic models. A predictive 3D multi-scale agent-based model (ABM) is proposed to investigate the role of cancer - stroma cell-to-cell interactions in multi-myeloma tumorigenesis. It includes: (a) Intracellular level: The intracellular signaling pathway features of myeloma initiating cells (MICs) and MM associated BMSCs may dominate biomechanically induced MM cancer cell phenotypes at intercellular level, cancer development and disease prognosis in the tissue level. (b) Intercellula level: Cell-to-cell interactions are the pivot chain linking intracellular level features of MIC an BMSC to intracellular biomechanical phenotype switch of MIC, BMSC, and progenitor cells (PCs) and MM. And (c) Tissue level: The cytokines secreted from MIC, PC and MM in the tissue level will regulate the proliferation and differentiation of MICs. In order to implement our goals, we set the specific aims: 1) Establish signaling pathway system using Modulated Factor Graph in regulating MICs and BMSCs by evaluating if modifying the CXCR4/SDF1 pathway changes the biomechanical properties of BMSCs and if these changes influence the proliferation and survival of MICs. 2) Establish MIC lineage model by developing quantitative cellular and cytokine assays to measure the amounts of different types of cells and secretary stimulatory/inhibitory cytokines using the well defined biological system. 3) Establish the predictive 3D MM growth model using Agent-based Model (ABM) by incorporating the biomechanical signaling pathways at intracellular level and the cell-cell interactions at intercellular level. The modeling system established will provide us a critical tool to see how we can manipulate the biomechanical interaction to interrupt the MIC development, which leads to the cure of myelomas.